NASA Perseverance Rover Executes First AI-Planned Martian Drive
NASA's Perseverance rover completes its first autonomous drive using AI vision-language models to plan routes. This breakthrough demonstration reduces reliance on manual human planning for Mars exploration.
NASA's Perseverance rover successfully completes its first drive planned entirely by artificial intelligence. During a December demonstration led by the Jet Propulsion Laboratory, engineers use vision-language models (VLMs) to create waypoints for the rover, a complex task that human planners traditionally handle manually. This milestone marks a significant shift in how NASA operates its robotic explorers on the Martian surface.
The AI system analyzes the exact same imagery and terrain data that human drivers rely on to generate these precise waypoints. Because Mars sits an average of 140 million miles away from Earth, a severe communication lag makes real-time remote operation impossible. Currently, human drivers must carefully sketch out routes with waypoints spaced no more than 330 feet apart to avoid hazards before beaming the plan through the Deep Space Network.
JPL's Rover Operations Center leads this initiative in collaboration with Anthropic, utilizing the company's Claude AI models to process the orbital imagery and slope data. NASA officials note that autonomous technologies like this allow missions to operate more efficiently, respond better to challenging terrain, and increase overall science returns as the distance from Earth grows during future deep space exploration.